Jump to content

Top Dawg Challenge


Recommended Posts

  • Replies 255
  • Created
  • Last Reply

Looks like 2 drives to me also..

yeh, and the best i can get with 2 drives is 228.. so lets not allow anyone with raptors to use their I/O scores.. i mean, raptors are obviously faster than maxtors.. just like nvidia raid is a faster raid.. so, if we strip down everyones I/Os, and video.. oh and AMD scores higher cpu scores here in the pit also.. lets get rid of that too.. and everyone with a 333 celeron can have a chance to win.. since about all that leaves is memeory scores. :blink: oh but wait.. AMD 64s have higher memory scores than the celerons do.. lets ditch that and make it a "who has the best pc name contest" :lol:
Link to comment
Share on other sites

yeh, and the best i can get with 2 drives is 228.. so lets not allow anyone with raptors to use their I/O scores.. i mean, raptors are obviously faster than maxtors.. just like nvidia raid is a faster raid.. so, if we strip down everyones I/Os, and video.. oh and AMD scores higher cpu scores here in the pit also.. lets get rid of that too.. and everyone with a 333 celeron can have a chance to win.. since about all that leaves is memeory scores. :blink: oh but wait.. AMD 64s have higher memory scores than the celerons do.. lets ditch that and make it a "who has the best pc name contest" :lol:

A valid point. However we still can't deny the obvious and wrong advantages of some parts. A FX 5200 scoring higher than the X800XTPE for instance..
Link to comment
Share on other sites

I just wished to voice a concern, as by no means am i an Expert. It was just a thought.

 

This may be a bottleneck issue created by 64 bit vs 32 bit computing, or it is a true bias issue which I felt should be heard.

 

I bow to the True Experts at the Pit.

Link to comment
Share on other sites

I do agree with Rmc that nvidia raid does get too many points, but these people with high I/O all have 4 drives (most have 4 raptors).

 

Pit video is an entirely different senario that nvidia fx cards and ti cards score way too high. :mrsgreen:

hmm..

 

ok, SIK_L_CELLS x800xt pe scores too high in 3dmark01.. lets omit his scores :rolleyes:

 

this whole world of benchmarking is all about scores.. you build your computer to get higher scores.. and for the pit test, that higher score, comes with nvidia, not ati.. and as shallow as it is, it's the main reason i got rid of my x800pro and got a 6800gt.. Aquamark on the other hand, seems to love ati more than nvidia.. and if i were building a computer to run aquamark, i'd get an ati.. but i built this one to run the pit test.. and to omit every advantage it has is kind of unfair.. a cap would be better than doing away with it..

 

like i said before, who has the best PC name?

if we took away everything that gives a certain product an advantage, and we all tested the same.. then why did i just spend $2500.00 upgrading, when i could have dug the old emachine out of the closet to achieve the same thing.. :blink:

Link to comment
Share on other sites

So, so far I agree that video scores got to go... we could go with the actual I/O score instead of the weighted score to help that problem...

I dont have a problem using AMark again with certain drivers and a minimum score say 50000, most of the 68's run around 70000-80000 without the null cheat.

Radeon 850 is gonna hit what 90000??? is that about right???

Sik and Sho... I need your AM scores

 

And yes... I switched back to my P4

Because I needed better rendering for Video production and I needed more HD space instead of using only 36gigs on my 80gig maxtors...

I'll put it back together soon, right now it's sitting in the closet in a box :mrsgreen:

Link to comment
Share on other sites

Just some thoughts here. I'm of the opinion that the Pit does not favor nVidia on video scores. The Pit measures 2D performance and ATI 2 D performance is not as good as nVidia. If ATI wanted to improve it they probably could. So to me it's what you want to measure not what's favored. Right now the highest 3D 03 scores on futuremark are with SLI boards and cards. The test isn't favoring anyone, it's the cards that are winning at the moment. By the time the next competition finishes it may be ATI's answer to SLI that is leading on that particular bench. If it were me I wouldn't concern myself with trying to even things out but I would try to get rid of cheats. 100,000 Aquamark using an X800XT PE isn't a cheat necessairly, but, someone could score 100,000 Aquamark with an X800XT PE using a cheat.

Deciding what we want to test in the next competition is a different question than deciding how we are going to limit cheating. IMO a Pitstop competition should include the Pit test. If we want to add a 3D test to that competiton then great. But why remove the 2D test. If you want to add an ATTO drive test great, but why remove the Pit I/O test. BTW the newest nf4 boards aren't scoreing as well or as consistantly as the nf 3 boards were on the Pit I/O. And if you used 4 drives on the nf3 boards you were limited to the amount of ocing you could do because there were no locks on the 1&2 sata ports. The hardest competition will test all of it. It will be up to the contestants to decide what they want to excell in and whether that will gain them enough points to take the competition.

Last competition we had to maintain the same hardware thoughout the competition and I think that was a great idea. I'd like to see the next competition have something like

1. Pit Test

2. ATTO TEST

3. 3D 03 TEST

4. 3D 01 TEST

5. Super Pi.

Some of the new SLI rigs will have 6800s and 6600s instead of the Ultras and they won't do so well in the 3D 01 tests. That will help the people with single card systems. Sorry for the book,, now I've gotten myself all excited. :P

Link to comment
Share on other sites

I'm scoring less in Aqua than I was but that's because I can't oc to the max yet because of Vcore and I'm using the plain 6800s. Here is one of my best AquaMarks with the X800Pro that's modded. http://arc.aquamark3.com/arc/arc_view.php?run=2096720691

SLI is hitting in the mid 80,000 with a moderate oc and plain 6800 cards.

 

Crossfire I'm betting he'll cap them or get rid of them but I think it's just like capping all the other benches. If you put a cap on them it's like saying the best doesn't win. What if 10 people hit the cap. What do you do then.

Link to comment
Share on other sites

i say we use the pit test, cap the video, leave the I/Os alone or hell, cap em at 200.. the folks with raid should have some advantage over the folks w/o it..

and use 3d05 and aquamark with the same drivers as from the last challenge.

That sounds fair! But how will you decide what percentages each test accounts for? Wouldnt adding them all up be a "middle of the road" compromise?
Link to comment
Share on other sites

That sounds fair! But how will you decide what percentages each test accounts for? Wouldnt adding them all up be a "middle of the road" compromise?

100% pit test, and 10% of aqua and 3dmark.. it would be the same thing if we figured it a 100% except it wont be any 200,000 point challenge scores :P
Link to comment
Share on other sites

So for example in the OC part the test you would expect to see something like:

 

Pit Test = 3000

Aq Mark= 9000

3D 03 = 1600

Total = 13,600.

That would be an example of my current scores without any adjustments or caps. A score calculated like that would give too much weight to AquaMark if you are trying to give them all equal wt. It would be weighted better to have Aqua be, score x .01 x 3.5=Total or in this case 90,000 X .01.X 3.5 = 3150

 

3D 03 could be SCore X .10 X 2= total or 16,000 X .10 X 2= 3200 Then the totals would look like.

Pit Test = 3000

Aq Mark= 3150

3D 03 = 3200

Total - 9350 For my current rig oc'd.

If you added PC Mark into the mix it would give some help to the Intel rigs which are going to suffer, but they are going to suffer all year long at everything.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Guest
This topic is now closed to further replies.

×
×
  • Create New...